It’s always great to be there at the beginning – when a nascent technology is forming and true believers gather to shape its future and share learnings. The Talkabot conference in Austin was one the first organized attempts (that I know of) to pull together platforms, developers, designers and VCs. Here are my five key take-aways from the conference:

1. Bots are a real thing. Granted that the conference had a self-selection bias. But the resources that tech companies are devoting to AI and non-AI powered bots is real. One of the prime motivations is that platforms and developers realize that no one is downloading mobile apps anymore.

From a larger philosophical perspective, “language as an interface” represents a step in the evolution of human interaction with technology. We talk with one another (using text or voice). Now we are able to “talk” with intelligent software – nothing to install or configure. No learning curve needed.

2. Several folks talked about chatbots being the next evolution of social. I think that is underselling the promise. Sure there are social functions facilitated by chatbots – the best example being Kip, a social shopping app. But chatbots are also work for solitary activities. Time will tell if this statement is hyperbolic, but Kik CEO, Ted Livingstone’s described Chat as the new browser and bots, the new websites. If you accept that, then we are at the dawn of Web 3.0 or are we at 4.0 now???

3. Chatbot discovery is an issue. Should chatbots live in an app store, such as the Slack app story? OR should they be found or shared just as one would a human? How can we avoid the same mistakes as mobile app stores that have become so crowded, that its virtually impossible to know that an app exists prior to discovering that you need said functionality?
An option outside of the store framework, is for friends to share apps (similar to WeChat), but that requires many people becoming the accidental “first-finders”. Perhaps deep-linking will resolve the discovery problem, but that is still a ways off.
The good news is that are several people grappling with this issue, including Botlist which curates and publishes chatbots.

4. Think past AI. Kik CEO, Ted Livingstone made a provocative statement where he asked developers to ignore the AI hype cycle. Great user experiences can be created by menu driven experiences and providing suggested responses. Per Ted, AI is being pushed by companies like Amazon, Google and Microsoft that don’t have popular chat platforms, but have lots of resources to devote to AI. The goal of chatbots is to create instant interactivity. To that end, Kik offered a coffee bot to conference attendees. I was able to scan a chatcode and send a request to the coffee bar. When my order was ready, Kik pinged me to go get my drink.

Others spoke of chatbots being the delivery vehicle of AI. We saw several demos where chatbots included buttons + NLP. Also saw several apps where humans are able to hijack (also known as handshake) a chatbot.

I tend to side with the latter. While simpler bots can be created by deep understanding of user expectations and actions, AI will build intelligence into the chatbots interaction. Machine learning will help the bot to improve over time. Today’s simple (even simplistic apps) will evolve into more human-like interactions.

5. No killer app yet. Several people mentioned that Chat Commerce is going to be the killer app that’ll mainstream chatbots. And clearly there is a food fight going on between the platforms to “own” the digital wallet. It was fascinating to see a demo of WeChat (which non-Chinese can’t access or build on). In China, WeChat is the internet, with payments integrated. WeChat bots can be share, invited into group conversations, show webviews, and process payments. Due to the more fragmented nature of our world, it’ll be interesting to see which platform wins the wallet war.

Additional tidbits:

• More than 90% of Slack paid users use apps and integrations
• More new bots launch on WeChat than websites launch on the web
• Just 2% of Slack apps are approved in the first round. I believe Kelasa was a second round approval – kudos to Sanat and team for making it so!
• Tex-Mex in Austin is as awesome as I remember